Thursday, January 4, 2018

A Technical Appendix to Postblogging Technology, November 1947, I: When America Gets a Cold, The Rest of the World Gets 70-Calibre Pneumonia"

Update: Put some words in.

At the head, I should point out that this appendix is up for the usual reason that I'm not going to get the next postblogging post up next week unless I can work on it this week. But, hey, the 3"/70 AA is morbid fun, and pneumonia might be going around this January. Relevance!

Can ships come down with pneumonia? Maybe there's a relief that can come in for a shift.
HMS Swiftsure is a cruiser that was scrapped in mid-modernisation in 1959/60 largely due to concerns about a proposed 3"/70 fit.
The postblogging series is still more than a year away from the final crisis of the Great Siege, the 30% devaluation of the pound on 19 September 1949. But, as Eric Groves helpfully reminds us, the devaluation was driven by the American recession, which reminded me of that old proverb about how the world gets pneumonia when America gets the cold; and we came across an earlier symptom of these ongoing problems this week, with the controversy over the reduction of the Home Fleet to a single cruiser and four destroyers. 

At one level, this reminds me of the Daily Mail's recent ginned-up outrage over the Royal Navy being reduced to "nineteen ships," in that it's complete bollocks. Submarines aren't ships, you see. Nowadays, the SSNs of the modern Royal Navy are the ships that keep the seas, and the nation's exclusive nuclear holocaust-related services providers (Yes, I have been catching up with my Laundry novels backlog over the holiday). Back in 1947, in the wake of the brief convertability episode in the summer, the Cabinet had forced cuts in the Estimates that reduced the surface elements of the Home Fleet to a single Dido-class cruiser and four Battles (with 1 battleship, 11 cruisers, 1 carrier, 24 destroyers and frigates and 5 submarines on foreign stations); but even though the cuts had been financial, their effective means of execution had been manpower cuts, and the fact that the Home Fleet was running twenty submarines probably tells us something about what was really going on. . . 

Which was only in very small part that the Admiralty was going full steam ahead with the submarine fleet.
HMS Alliance is a historic exhibit. Good Lord, that's a big sonar. [And this is opinionated, but pretty good.] By Keith Edkins - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=759320
Hydrogen peroxide, hence "Blonde submarines"
It's a hilarious Late-Forties joke!
The fleet included refitted "Ts," eventually with American high-density batteries, although not yet in 1947, the brand-new "A-" class,HMS Meteorite, trying not to explode, some lucky boat also trying not to explode as it field-tested the "Zombie" hydrogen peroxide powered torpedo, and "S"-boats converted as high speed targets, including HMS Scotsman, in refit from October, to emerge in 1948 with a silent, experimental electric motor that would give ASW trial forces fits, although the Wiki article doesn't even mention it! All in all, I'm glad that I'm not trying to post on the transformation of the Royal Navy submarine fleet in 1947! I did spend a bit of the morning looking for a source that would discuss it, and have linked to Peter Hennessey and James Jinks' guest blog post at RN Subs, above. The key point, buried a bit in all of that verbiage is "refit." It will come up again.



The other bit ticket item was the Royal Navy's ongoing effort to transition to a carrier-aviation based force centred surface force, currently scheduled to complete in 2023. Seventy-eight years seems likely somewhat leisurely progress to me, but what do I know? Anyway, as of the fall of 1947, Triumph is down in the Mediterranean, flying 28 Seafires and Fireflies; all of the exciting news about jets landing on carriers is presumably being generated by ships operating strictly on a trials status. What's going on? I have the data!

Since there's a risk that the empirical will swamp overarching, theories, I'm going to summarise what the Colossus-class light fleet carriers were doing in the fall of 1947.
i) Coloussus had been turned over to the French on 6 August 1946, and would have a long and illustrious career as Arromanches.
ii) Edgar had been designated for conversion to a maintenance ship soon after launch, and renamed Perseus. Placed in reserve at Rosyth in 1946, Perseus was selected as trials ship for the first steam catapult at about the same time, but the installation would have to wait for a scheduled refit in 1950, although my source (David Hobbs) is silent on the question of whether the catapult was ready for installation earlier.  
iii) Glory had paid off on 18 August 1947 after returning from the Pacific. Currently being used as an accommodations ship for Royal Australian Navy personnel waiting for Terrible/Sydney to complete. It would commence a major refit in the summer. 
iv) Mars was designated as a maintenance carrier in the summer of 1944, and renamed Pioneer. After postwar service ferrying troops, Pioneer was paid off in the summer of 1947 into the unmaintained reserve, surplus to requirements, and was broken up in 1952.
v) Ocean was in service with the Mediterranean Fleet.
vi) Theseus was steaming back from the Pacific for a refit that would commence in December. So describing it as not in service is both accurate and, in terms of resources committed, somewhat misleading.
vii) Triumph was in service with the Mediterranean Fleet. 
viii) Venerable had been declared surplus to requirements in the fall, but was hanging around at Devonport, because the sale to the Dutch, effective in the spring, was already under negotiation.
ix) Vengeance was steaming to the Pacific to replace Theseus upon arrival, so see my comments for that ship.
x) Warrior was serving with the RCN until the Majestic-class Magnificent was ready, after which it would return to Britain in the spring of 1948 to serve as the flexible-decks trials ship after a refit.

I have to admit to being a bit disappointed that Eric Groves' claim that there was only one carrier in service in the crisis of 1947 appears to be not just misleading, but off by one ship, and it's likely that I'm misunderstanding him about Ocean and Triumph. Counting out both the carrier coming back from the Pacific and the one headed out there seems a bit dodgier. The scale of the crisis of 1947, serious enough, is being overstated.


. . . Which brings me to by far the least important aspect of the postwar fleet refit/modernisation/building programme, the cruiser force --the one that, nevertheless, manages to suck all the oxygen out of the room, because "information technology"/"automation"/"robots taking our jobs," etc. 
The story here could easily start with the Hawkins-class cruisers of WWI, of which Vindictive, converted into an earlier aircraft carrier, is seen above. Britain, the United States, and the Japanese Army emerged from the world war keen on the idea of a naval arms limitation treaty that would prevent another ruinous dreadnought race. Each had their own reasons, and it doesn't take a PhD in political science to see why the Japanese Army and Navy might be at odds, here. Unfortunately, the British talked themselves into needing a vast fleet of cruisers to police the world's trade lanes, a logic that was unconvincing to both America and Japan on the grounds that they needed as large a fleet as the British, for dick-measuring purposes, and didn'need a vast fleet of commerce-protecting cruisers, for obvious reasons. Here things might have ended had the British not decided to build a giant cruiser, which, of course, led the Americans and Japanese to agree that while they didn't need cruisers, they did need giant cruisers.

Because dick-measuring. Or, more likely, because they saw a sore point that they could press to get something out of the British. The eventual compromise was that the British could build a lot of cruisers, but so could the Americans and Japanese, but while they could be huge, they had to be armed with 6" guns only. This, of course,makes complete sense when you look at all the factors under consideration and [snore]...

So the Japanese promptly saw a technicality, which was that you could make a giant-sized cruiser and just cram 6" guns into the hull, and what about that, hunh, hunh?
6" guns, but fifteen of them, equals Japanese more manly than public-school going Britons who like to be paddled by upper form boys, if you know what I mean, and I think you do.


A provocation not to be born, naturally leading to
HMS Southampton may or may not be the lead ship of the "Town-" class cruisers, because the naming conventions are a mess. 


At this point, sanity very briefly intervened, leading to a cruiser designed to do the job, as opposed to competing in the all-important dick-measuring contest:
Sure, it's cheap, efficient, and designed for the vitally important role of fleet anti-aircraft defence, but its guns are too small.
Move forward to 1942, towards the end of Dudley Pound's tenure as First Sea Lord, and discussion about a new cruiser class to follow on the wartime "Town" class permutations. (Of which Swiftsure and its sisters, and the Tigers, the last cruisers actually built for the Royal Navy, inasmuch as we don't call the "Counties" and later "destroyers" "cruisers," because reasons, were members.) Pound not being insecure about his manhood, a Dido replacement sailed through committee.

Then Pound died and was replaced by Andrew Cunningham, and the small, 28 knot, 5.25" gun-armed Dido replacement became the 15,000 ton Neptune,  with 12 6" guns, just like the "Towns." In fairness to Cunningham's Board, the gun cruisers of this era tended to grow. USS Worcester was 14,700t standard displacement, Newport News was a grotesque 21,000t, and even the Sverdlov, built, as Zhukov cruelly but accurately put it, to be scuttled on the outbreak of war, after doing its strategic duty by forcing the sea powers to spend on their navies instead of their armies, weighed in at 13,600t. 
Given that Sverdlov motivated the Blackburn Buccaneer, I'm calling it a win-win.
That being said, the Neptunes showed a suspect strand of giganticism, in that in some ways it was the gun that it was to carry that drove the ship design, and not the other way round, and the gun was something else. The QF 6" Gun Mark N5 was a 50-calibre weapon firing an oversized, 126lb HE shell, with a "hydraulically operated" breech. That's code for "automatic gun." The same concept was seen on Worcester and Newport News, and in the Mark 26 mounting chosen for the Neptunes, was supposed to elevate to 85 degrees and fire at 10--12 rounds per minute. Controlled by two barrage directors in high angle fire, the 6" guns would have been an effective anti-aircraft "screen" weapon protecting the fleet against oncoming Kamikazes and other relevant foes, in case World War II happened again. In the more likely event that WWII didn't happen again, and the threat was cruded anti-ship missiles and glide bombs, the Neptunes also carried the same 4.5" dual-purpose mountings chosen for the Daring-class destroyers (speaking of the Cunningham Board's giganticism).

At this point, reality, of which I will say more in a bit, began to set in, the Neptunes began to recede, and the QF 6" Gun Mark N5 was repurposed as a proposed weapon for the Tiger-subclass of the "Towns." Moreover, the whole idea of new and exciting guns for (not-very) old cruisers was reaching something like maniacal proportions. The Minotaurs, immediately predecessors to the Tigers, were not to get the new 6" gun, but they were to get another new gun, the  3"/70, which is where I came in. Since this is going to be the key technical issue in this post, I will point the reader in the direction of H. W. Pout's analysis of weapon direction in the Royal Navy, which uses the ballistic performance of the 3"/70 as a starting point in analysing the error contributions of various elements of the direction system. The publisher has very kindly posted the first page of this study online. It's admittedly a highly indirect introduction to this weapon system, but it's not like there's much better, out there. 

Swiftsure, fresh from shooting at Communists in Korea, went into drydock in February of 1957 for a major refit that would install the 3"/70. the Wikipedia article, which is all I have, seems to have suffered from some rather choppy editing, but is still something of an introduction to the slowly-evolving scandal of the protracted and over-budget refit, from which Swiftsure was eventually released to the breaker. It's probably an oversimplification to say that it was caused by the teething problems of the 3"/70, but this somewhat utopian technology can't have helped. The Tigers, by way of contrast, marched through their refits, eventually emerging as "helicopter cruisers," still with 4 N5s and 4 3"/70s in two twin mountings, each.

Tiger and Blake made it almost all the way to the Falklands War, and were actually recalled from reserve and under refit for dispatch after the main task force, until the Admiralty belatedly woke up to the fact that it was proposing to sending 1770 crew to a combat zone, protected only by AA guns and with very little more to do than serve as forward Harrier pickets, the Argentine submarine force having declined to commit suicide on an entire navy's worth of nearly-entirely ASW dedicated assets.

Chickens.

Speaking of being reasonably scared of scary things, the 3"/70 dates back to the Kamikaze crisis of 1945. In a rousing round of general incriminations, it was agreed that the 20mm Oerlikon wasn't pulling its weight; that the 40mm Bofors shell was too small for a proximity fuze, and that the 5"/38 was too slow, both in training speed and muzzle velocity, to engage kamikaze targets. The closest weapon to satisfactory was a 3"/50 mounted on the destroyer escort force, but it was manually loaded, and this was deemed to give it too low a rate of fire. Under the circumstances it was perhaps understandable that there would be a compensating overreach, and the design that emerged from the Anglo-American think tank was an automatically-operated 3" gun of 70 calibres length.

Pause and let that sink in for a moment. We are talking about a 1,200kg barrel of 210" (almost 5 meters), which works out to a weight of something like 200 grams/centimeter, a number that I am guesstimating in the service of trying to get a grip on the torque required to drive that muzzle at its 50 degree per second training rate/ 35 degree per second elevating rate at anywhere between -15 degrees and 90 degrees elevation. This not being grandiose enough, the designers aimed for a muzzle velocity of 3500 ft/s, which is faster than the sabot ammunition fired by the German 75/L70 antitank gun, only firing a more-than-full-sized 6.8kg shell (16.4kg complete round). While no information on the propellant composition or barrel design is available to me, I would eat my hat if we are not by now in the postwar era of RDX-doped, triple-base propellants and carbonitrided, chromed barrels, as barrel lifetime is an expected 2000 rounds. 

It is perhaps not surprising that this very ambitious weapon concept took to the sea on USS Carpentera refitted wartime Gearing class whose "modifications emphasized electronic equipment and anti-submarine warfare (ASW) weaponry over the standard destroyer anti-aircraft and torpedo armament." A small number of other USN high-concept ASW escorts of the 1950s got the same weapon, and, from there, it departed into obscurity with the Tigers. 

There are three, intertwined stories here: the first is the crisis of 1947; the second is the anti-kamikaze guns of the postwar era; the third is the underperforming postwar cruisers of the Royal, and every other navy; the fourth, which sort of slipped in above, is cybernetics. There's a reason that I let cybernetics slip into the story, though. I don't know what "cybernetics" is. I'm hoping that the reviews of Norbert Wiener's Cybernetics will tell me, once it comes out in a few months (retrospectively from the point of view of the postblogger). 

When Ant-Man premiered in 1962,cybernetics was the jam, so they said that he could talk to ants because "cybernetics." Marvel seems to have dropped that angle, but I forgive them because the movie was funny. 

Okay, that's facetious, and I've already done more than my share of that with my scathing reference to my beloved Tom Kuhn's all-too fertile coining of "paradigm."  It's just that it has always been easier to say what cybernetics was about, than what it was, and considering that there is nothing more cybernetic-y than an automatic anti-aircraft gun intended for shooting kamikazes down, that's probably the place to start.

The problem of AA fire control is very simple. You have an aircraft: you need to shoot it down. That means that you need to solve for the vector of (a) shell(s) fired now that intercepts the plane in minimum time. For that, all you really need is the target's current vector, and some way of constraining its future acceleration to be within the dangerous space of your shell(s). To get the vector, you need its current location, which can in theory be given by a rangefinder (optical or radar), and some device that returns the angle at which the rangefinder must be driven to "lock on" to the target. Long before anyone even tried to resolve this problem in detail, it occurred to someone, possibly Arthur Pollen and possibly someone even before him, that the rate at which the rangefinder must be driven to keep up with the target is the radial velocity of the target. In practice, that insight only became useful with the invention of radar, which can actually lock on to the target.

This last bit is fascinating. Any change in current strength will produce an equal and opposite response, the so-called "back emf." So, in theory, at least, if you have a radar centred on a target, in the sense that the return signal strength is at a local maximum, any movement of the target away from that maximum will produce an induced current. Amplifying that current and applying it to the radar's steering motor will produce a feedback that tends to turn the radar back to the target. I, personallyl, am not a radar engineer, and have no idea whether that is how "lock on" works in practice, but the idea of an automatic feedback mechanism that keeps the radar, and hence the gun it is directing, on target, is fascinating. It is as though the enemy aircraft is steering or governing (Greek, "cyber" and "netics," says Wikipedia and Norbert Wiener) the gun. 

If, at this point, you have noticed that Royal Navy fire control solutions continued to use regenerative methods (that's the polite way to say "guessing") into the 1950s, then, well, congratulations, you're ready to write a book or two about how stupid the Admiralty was. [1,2,3] It's this library of books about stupid admirals that I referred to above as "sucking all the oxygen from the room," as by the time the recriminations over Swiftsure were over, it was settled fact that British industry/society/the armed forces literally could not do anything right, a point upon which Maggie Thatcher was eventually able to capitalise, and here we are today.
Everything American is Better.


That is not, however, or should not be, the issue today. Instead, I just want to focus on one of Pout's error sources, the mechanical "backlash" inherent to any automatically driven system. Backlash will be familiar to anyone who has oversteered an old-time automatic transmission, or, actually, any number of other mechanical systems, now that I think about it. It needs to be damped out, and the more torque you're pushing through the system, the harder it is to damp. Which probably and accurately describes what was going on with the 3"/70 in the late Fifties, when this wonder technology that automatically feeds ammunition from the magazine into the gun and hydraulically rams it home, firing it to on instruction of the director even as the muzzle of the gun is wobbling about at a rate of, as mentioned, 50 degrees/ in the horizontal and 35 degrees/second in the vertical. 

It's this, I think, that made cybernetics so exciting.
I ain't letting it go, on account of Evangeline Lilly. 

It's a way of understanding how the emerging field of electronics, and the apparently emerging field of hydraulics [pdf --it's not history, but I think this is one of those history of technology cases where the history has yet to be written] are going to mesh with the problem of automating everyday work, before we decided that we'd think about it as "information technology," instead.  So that brings me to the long-promised bit where I get "real," and it is, of course, in connection with the naval readiness crisis of the fall of 1947.

I have, above, been snippy about Eric Groves, but he does, in fact, give a solid account of a now-forgotten moment in British history that certainly happened. There was, in fact, a point in the fall of 1947, at which cuts in the Naval Estimates brought the Home Fleet, a force, of theoretically astronomical strength, down to a single cruiser and four destroyers; and the fact that we would actually have to pile on submarines and minesweepers and trials ships to get an accurate assessment of the Home Fleet's Numbers Borne doesn't change that. 

What's at issue here is not so much the facts as the interpretation, and the interpretation is . . . vexed. We want to look at this and issue a lapidary judgement about how the Admiralty's ambitions were excessive in a postwar period of financial stringency and austerity, national economy overstrained by war, urgent need for economies, etc., etc. That was, after all, the story of 1919, and the story of many a cancelled warship or other weapons system development programme of later times. Swiftsure is the header image for a reason!

And yet. . . 
By Barry Lewis - Portrait Of A Destroyer, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=27495112
At the same time that Swiftsure was going through its agonies, the Royal Navy was quietly adding 8 "County"-class destroyers to its strength, and while "destroyer" sounds small and innocuous, we are talking about a 6,200 t displacement (so, larger than a Dido), and a design built around the Seaslug missile, a far longer delayed, more expensive and technically ambitious weapon system than the 3"/70. In fact, the only difference between the "Counties" and the Minotaurs that really jumps out is that the missile-armed, steam-and-gas powered Counties" had a complement of 471, compared with the Swiftsure's pre-refit 867, or the Tiger's post refit 885. 

Not exactly coincidentally, back in 1948, when Vanguard goes out of commission for a refit, the Admiralty will decline to reactivate HMS Howe to replace it. As much as the Admiralty continues to value battleships in the postwar era, scraping up enough engine room ratings to commission the Howe will require immobilising a cruiser and four frigates. Even in a shooting war, there was no doubt alternative was more useful. 

The reality was that the services proved well able to keep the Government from pushing the Estimates down to a constantly argued-for 700 million pounds. The money situation was not critical. Few British governments have been more flush, in money terms, than postwar Labour. Plenty of other things were short, most notably US dollars, but the crisis that immobilised the Royal Navy had more to do with the fact that it couldn't get the re-enlistment rate for stoker ratings above 20%, compared with a prewar 60%. This was a problem that could be solved with money, at least, hopefully, by paying stokers more, but it was caused by the low civilian unemployment rate, and not austerity.

However, it was also caused by the export drive. After WWI, and to some extent in other austerity crises, the argument was that shipbuilding was needed to keep the vital defence industry going. In this crisis, the argument was that naval shipbuilding would get in the way of commercial shipbuilding for exports. So the reason that the Navy is going slow with all of its refits and upgrades is not because of lack of money, but because of lack of slips. There was too much work! Maggie, fortunately, will fix that, not least because the Argentinians couldn't wait a year to have their victorious foreign war.

This has been your weekly reminder that successful automation is an epiphenomena of a hot labour market. I guess the interesting question is whether failed automation is an epiphenomena of a weak labour market. (He types as, at work, the Computer-Assisted-Ordering software some shyster sold to Sobey's over-orders store-brand boxed spinach and under-orders the premium brands, again and again and again.)

You'd think that it would notice that we keep running out of this stuff and throwing out the other stuff. but that's not actually how "deep learning" works. It's kind of like the way that trying to apply tachimetry at sea produces a greater error than regeneration.

2 comments:

  1. Pound not being insecure about his manhood, a Dido replacement sailed through committee.

    Pound being asleep in committee, quite possibly.

    ReplyDelete
  2. Potato, potahto. It's one way of not being insecure!

    ReplyDelete